News Heat Trend
Project Overview
Frontier is a Crypto & DeFi, NFT wallet where users can send, store, and invest in over 4,000 crypto assets. Users can also earn passive income on crypto by staking or supplying assets in DeFi apps, and explore web 3.0 from a single platform.
Alliance Partner Discusses a16z's Shift Toward Frontier Tech Investment: AI Infrastructure Development Mirrors Crypto's Past, This Shift May Spur Crypto Talent to Build Frontier Tech
Alliance General Partner Imran recently posted on X, sharing his views on a16z crypto's shift toward frontier technology sectors. He noted that a16z provided early endorsement for the crypto industry when Chris Dixon led Coinbase's investment in 2013; Marc Andreessen's 2014 New York Times op-ed "Why Bitcoin Matters" was a key factor that prompted him to transition into the Bitcoin industry. At the time, very few institutions and funds were willing to publicly support crypto, and a16z and USV were the two that took on brand risk.Imran believes that the current development of AI infrastructure has benefited considerably from Bitcoin mining infrastructure and talent flowing from crypto to AI. Shifting toward frontier tech such as robotics, biology, AI, energy, and compute means that the talent and capabilities accumulated in the crypto industry over the past decade can be used to build the entire frontier tech sector. Citing data centers as an example, he noted that public and government resistance to data centers is growing, much like earlier resistance to crypto—yet today nearly all traditional financial institutions have embraced crypto to some extent.
Dragonfly Managing Partner: Onchain Pre-IPO Derivatives Show AI Lab Valuations Not Significantly Impacted by "Slowing the Frontier Race"
Dragonfly managing partner Haseeb Qureshi stated that price movements in onchain Pre-IPO derivatives reflect interesting market expectations for AI lab valuations. He noted that Jacob Coxon's earlier pessimistic remarks about AI risk actually pushed related lab valuations higher, arguing that "scaring everyone seems to be good for business."Haseeb said that after Anthropic CEO Dario Amodei proposed "slowing the frontier race," related valuations were not significantly affected, which is not surprising, since reducing the model race means training spending declines, thereby lowering major capital expenditures. However, after David Sacks publicly criticized Dario, Anthropic's valuation was largely unaffected, while OpenAI's valuation declined, and he raised questions about the reason for this.
Bank of England Governor: Frontier AI Models Threaten Global Financial Stability
According to CNBC, Andrew Bailey, Governor of the Bank of England and Chairman of the Financial Stability Board (FSB), wrote to G20 finance ministers and central bank governors on August 31 to warn that frontier AI models are exhibiting increasingly sophisticated autonomy and threat capabilities, which could fundamentally alter the speed, scale, and economics of cyberattacks, potentially triggering disorderly adjustments across global financial markets. Bailey noted that highly concentrated third-party service providers will become critical nodes of systemic risk, emphasizing that most countries currently lack regulatory frameworks for the development and deployment of frontier AI. He also flagged vulnerabilities in sovereign debt markets, the growing trend of leverage in equity markets, and the overvaluation of AI-related assets, urging governments to accelerate the refinement of relevant safety mechanisms.
Serenity: The AI Supply Chain Boom Is Far From Over, With Storage, Packaging, and Power Sectors Poised for Long-Term Demand Surge
Odaily News "White-Haired Stock God" Serenity shared insights on the AI industry chain on the X platform, noting that AI infrastructure demand is driving multiple sectors—including storage, advanced packaging, computing power financing, optical communications, power supply, and electronic components—into a long-term expansion cycle. The AI supply chain remains in a phase of rapid growth.In the storage sector, Serenity cited UBS forecasts indicating that traditional DRAM manufacturers (such as Micron) could see gross margins reach an unprecedented 95% by 2027, potentially even surpassing the gross margin levels of HBM products. Additionally, SanDisk's long-term agreements already cover approximately two-thirds of its 2028 production capacity, with minimum contracted revenue reaching $93 billion. Given its current market cap of around $239 billion, this suggests its future revenue targets could persist for years, making it difficult to simply classify the company as a traditional cyclical stock.On the cloud computing infrastructure front, CoreWeave has signed agreements to use Nvidia A100 GPUs through 2029. This is a positive development for emerging cloud computing companies such as Nebius and Iren, and it also weakens some investors' bearish thesis centered on the rapid depreciation of older GPUs.AI model companies are also continuing to grow at a pace that exceeds expectations. Frontier AI labs are still maintaining extremely rapid growth rates, and a slowdown in growth would actually be a cause for concern. The market projects that Anthropic's 2028 revenue could reach $190 billion to $200 billion.However, advanced packaging and semiconductor infrastructure remain core bottlenecks. The head of advanced packaging at TSMC has stated that in the coming years, the industry may face not only memory shortages but also tight supply of ABF substrates.Serenity concluded that the AI infrastructure supply chain is continuously expanding. From GPUs, storage, and advanced packaging to power, optical communications, and electronic components, every segment is showing a long-term demand growth trend. The AI supply chain is still in a high-speed development stage.
Framework Ventures Raises $400 Million Fund Targeting Tokenization, Stablecoins, and Frontier Technologies
San Francisco-based venture capital firm Framework Ventures has raised a $400 million fund, planning to invest in the intersection of tokenization, stablecoins, and frontier technologies. Framework Ventures co-founder Michael Anderson stated that the firm believes blockchain, through tokenization, will become the financing layer for AI computing power, robotics, and energy infrastructure. The crypto industry has shifted from serving crypto users to solving capital formation problems for real-world industries. (CoinDesk)
a16z Crypto rebrands as a16z Frontier, expanding investment scope from crypto to frontier tech startups
a16z co-founder and crypto business lead Chris Dixon announced that a16z Crypto will be renamed a16z Frontier, expanding its investment scope from crypto startups to a broader range of frontier technology sectors, with a focus on supporting companies developing new technologies, creating new markets, and potentially facing longer commercialization cycles. Dixon stated that this adjustment represents business expansion rather than an exit from the crypto space; the fund will continue to invest in crypto and related projects and promote sound crypto regulatory policies. The new fund will also focus on AI and the new products and markets it spawns, including autonomous machines, hardware-software integrated startups, and new types of networks connecting humans and agents.Dixon believes that blockchain will play an important role in this round of technological transformation, that stablecoins are becoming a common tool for fintech companies and traditional institutions, and that blockchain is also expected to support machine-to-machine transactions and provide native economic capabilities for new types of networks. The fund will be jointly led by Dixon, Marc Andreessen, Ben Horowitz, along with Eddy Lazzarin and Guy Wuollet.
David Sacks: Leading U.S. Frontier AI Labs Sign White House SI Agreement, Will Introduce Internal Controls and External Audits
Odaily reports: David Sacks stated that heads of major U.S. chip, data center, and frontier model companies have held a meeting on the development of "superintelligence," and frontier AI lab leaders have signed the White House Superintelligence Agreement. The agreement is voluntary in nature and aims to strengthen AI safety governance through industry self-regulation.Sacks stated that signatories will assume responsibility for safe product development and implement new internal controls and independent external audit mechanisms. He believes that compared to waiting for international-level regulatory agreements, this model can establish AI safety constraints more quickly while maintaining U.S. development in related technology fields. The above assessment of policy effectiveness represents Sacks' personal views.
Institutions: AI Giants Worry They're Moving Too Fast, But Most of Their Customers Are Still Far From the "Frontier"
Over the past year, the AI field has experienced explosive growth, and this pace seems to have left frontier AI developers struggling to keep up in a short period of time. Now, these companies are beginning to call for a slower pace of research and stronger regulation and oversight.However, the AI safety crises that have previously been brought to light are unlikely to similarly impact the customers of AI giants like Anthropic and OpenAI. Enterprises typically deploy models that have already been commercialized and whose capabilities are relatively well understood by the market. These models may not be at the absolute cutting edge of AI, but their capabilities are already powerful enough to handle the various tasks enterprises assign to them.Stephen Messer, co-founder of Collective, said that most enterprises are not using the most cutting-edge models. They may even still be using versions from two years ago, and for the work they are doing, these models are actually more than sufficient. Some users don't even really know how to effectively prompt the models. Therefore, if the pace of AI development slows down, it won't have much of an impact on them. (WSJ)
Tech investor Jason: AI stocks will drop more than 10% in early trading on Monday
Odaily reports: Well-known tech investor Jason posted on X: "AI stocks will drop more than 10% in early trading on Monday," adding that Amodei's blog post "single-handedly dismantled the AI trade."Anthropic CEO Dario Amodei previously published a blog post with a title that directly stated: "We Must Pace the Frontier." Within hours of the article's publication, SpaceX CEO Elon Musk posted on X, saying "Dario is right."Early market indicators show that following the "AI slowdown" news, related assets on the HyperliquidX platform fell in response, with OpenAI and Anthropic dropping 7% and 2.8% respectively. Multiple market observers and investors have issued warnings on social media, concerned that the news will cause a significant impact on AI-related stocks such as Nvidia, Broadcom, and AMD on Monday.Executives at Google DeepMind and Meta have not yet publicly commented on this weekend's discussions. Divisions within the industry also cannot be overlooked. Critics argue that leading AI companies have self-interested motives in emphasizing existential risks—expensive safety and compliance frameworks could shut out smaller competitors, thereby consolidating the market position of existing giants.
Wyoming Expands Partnership with Chainlink to Introduce On-Chain Reserve Verification for FRNT Stablecoin
According to The Block, the Wyoming Stable Token Committee announced the adoption of Chainlink Proof of Reserve to provide near-real-time on-chain reserve validation for its official stablecoin, Frontier Stable Token (FRNT). The Network Firm will audit FRNT reserves in accordance with AICPA standards, while Chainlink will post verification data on-chain in real time to bridge information gaps between reporting cycles. Previously, Wyoming fully migrated FRNT from LayerZero to Chainlink CCIP last month as its sole cross-chain infrastructure. The committee is also advancing the Chainlink Proof of Reserve Secure Mint feature, which requires verifying that reserves do not fall below FRNT's total supply before minting new tokens to prevent infinite minting attacks. FRNT launched in January this year and is the United States' first government-issued stable token, backed by U.S. dollars and short-term U.S. Treasuries.
Bank of England Governor: Frontier AI Models Threaten Global Financial Stability
According to CNBC, Andrew Bailey, Governor of the Bank of England and Chairman of the Financial Stability Board (FSB), wrote to G20 finance ministers and central bank governors on August 31 to warn that frontier AI models are exhibiting increasingly sophisticated autonomy and threat capabilities, which could fundamentally alter the speed, scale, and economics of cyberattacks, potentially triggering disorderly adjustments across global financial markets. Bailey noted that highly concentrated third-party service providers will become critical nodes of systemic risk, emphasizing that most countries currently lack regulatory frameworks for the development and deployment of frontier AI. He also flagged vulnerabilities in sovereign debt markets, the growing trend of leverage in equity markets, and the overvaluation of AI-related assets, urging governments to accelerate the refinement of relevant safety mechanisms.
Formerly lost $23.75 million, Ostium's $15 million loan dispute goes to court today
According to court documents, a federal court in New York will hold a hearing today in the case brought by ESS Frontier and Hong Kong lender Minghui Zheng against Ostium, involving a $15 million loan dispute. The two parties are asking the judge to freeze Ostium's assets and refer the dispute to arbitration; Ostium previously lost $23.75 million in an exploit in July.
Wyoming Expands Partnership with Chainlink to Introduce On-Chain Reserve Verification for FRNT Stablecoin
According to The Block, the Wyoming Stable Token Committee announced the adoption of Chainlink Proof of Reserve to provide near-real-time on-chain reserve validation for its official stablecoin, Frontier Stable Token (FRNT). The Network Firm will audit FRNT reserves in accordance with AICPA standards, while Chainlink will post verification data on-chain in real time to bridge information gaps between reporting cycles. Previously, Wyoming fully migrated FRNT from LayerZero to Chainlink CCIP last month as its sole cross-chain infrastructure. The committee is also advancing the Chainlink Proof of Reserve Secure Mint feature, which requires verifying that reserves do not fall below FRNT's total supply before minting new tokens to prevent infinite minting attacks. FRNT launched in January this year and is the United States' first government-issued stable token, backed by U.S. dollars and short-term U.S. Treasuries.
Bank of England Governor: Frontier AI Models Threaten Global Financial Stability
According to CNBC, Andrew Bailey, Governor of the Bank of England and Chairman of the Financial Stability Board (FSB), wrote to G20 finance ministers and central bank governors on August 31 to warn that frontier AI models are exhibiting increasingly sophisticated autonomy and threat capabilities, which could fundamentally alter the speed, scale, and economics of cyberattacks, potentially triggering disorderly adjustments across global financial markets. Bailey noted that highly concentrated third-party service providers will become critical nodes of systemic risk, emphasizing that most countries currently lack regulatory frameworks for the development and deployment of frontier AI. He also flagged vulnerabilities in sovereign debt markets, the growing trend of leverage in equity markets, and the overvaluation of AI-related assets, urging governments to accelerate the refinement of relevant safety mechanisms.
Crypto Companies Send Joint Letter to AI Labs, Urging Access to Frontier Models for Bitcoin Developers
据 Cointelegraph 报道，比特币政策研究所（BPI）联合 Anchorage Digital、BitGo、Bitwise、Blockstream、Kraken、Ledger、MARA、Trezor 等多家加密机构，发布公开信敦促各大前沿 AI 实验室为比特币及开源软件开发者建立或扩展可信访问计划。 信中指出，Bitcoin Core 等开源维护者目前缺乏对 AI 实验室网络安全程序的访问渠道，被迫依赖能力较弱的开源模型，而比特币网络当前保护着逾 1 万亿美元资产，任何开源基础设施漏洞均可能危及用户毕生积蓄。BPI 同时披露，已收到多份报告显示包括潜在境外势力在内的复杂攻击者正借助先进 AI 能力持续发动攻击。
1178 employees from top AI companies sign joint statement calling for international cooperation to control the pace of frontier AI development
1,178 employees from leading AI companies have jointly released a statement titled "Pacing the Frontier," urging the U.S. government to support international cooperation in developing technologies and governance tools to proactively regulate the pace of frontier AI development.The statement says that global leading AI companies believe the future may be approaching the point where automated AI research becomes achievable. Although it is impossible to predict exactly how this will accelerate AI progress, there is a risk that AI capabilities could rapidly surpass human understanding and control.The signatories stated that the industry, governments, and society need to retain the option to "buy time" to address emerging risks, improve safety measures, and strengthen regulation. However, due to competitive pressures faced by both companies and nations, the world currently lacks governance mechanisms capable of proactively controlling the pace of frontier AI development.The statement notes that frontier AI agents are already capable of discovering and exploiting real software vulnerabilities, and could be used for large-scale cyberattacks in the absence of safety measures.Employees from multiple organizations including OpenAI, Anthropic, Google, and Meta have signed the statement, and believe there is a need to establish international coordination mechanisms to reduce potential risks while advancing AI development.
Anthropic's Restricted Spillover Effect Emerges: Asian AI Companies Simultaneously Launch Rival Frontier Models
as Anthropic faces export restrictions limiting the global availability of its advanced models, multiple Asian AI companies are accelerating efforts to fill the market gap. Chinese cybersecurity firm 360 Security Technology has reportedly launched an AI tool called "Tulongfeng," claiming it can directly compete with Anthropic's high-end model "Mythos." Meanwhile, its more restricted version, "Fable 5," also falls within the scope of relevant export controls.In the same week, Japanese AI startup Sakana AI released a new model named "Fugu," taken from the Japanese word for pufferfish. It is positioned as a frontier model designed for agents. The company stated that the model's capabilities are comparable to Fable 5 and Mythos Preview, and it supports coordinating multi-model calls via API to enable agent orchestration.Sakana AI emphasized that the timing of this release and the U.S. export restrictions are "purely coincidental," but the product's official website still clearly promotes "providing frontier capabilities without the risk of export controls." Company co-founder David Ha stated that future AI development will shift from competition among single large models to "model orchestration systems," adding that "access can disappear at any time, and distributed intelligence is a realistic hedge against the risks of centralization."On the other hand, Chinese 360 founder Zhou Hongyi views AI vulnerability detection capabilities as a "national strategic asset" and warns of the so-called "one-way transparency" risk, where certain entities may monopolize advanced security capabilities.According to reports, the U.S. export restrictions on Anthropic's advanced models have been in place for about two weeks. Against this backdrop, Asian manufacturers are accelerating the launch of local alternatives. Although some companies still emphasize the importance of American models in the Asian market, the trend of differentiation within the regional AI ecosystem has begun to emerge. (TechCrunch)
Sky Q3 net revenue reaches $37.19 million, posting positive growth for six consecutive quarters.
According to disclosures from Sky Ecosystem Insights, the Q3 2026 report released by the Sky Frontier Foundation shows that Sky's net revenue reached $37.19 million in the quarter, representing a year-on-year increase of 11.8% and marking six consecutive quarters of positive growth. The net profit margin rose from 31% in the previous quarter to 37.5%. During the same period, total protocol revenue amounted to $99.29 million, up 14.2% year-on-year, while total expenses decreased by 16% year-on-year to $62.10 million. At quarter-end, the protocol's total collateralized assets reached $11.41 billion, with sUSDS supply standing at $4.52 billion, up 118% year-on-year.
Alliance Partner Discusses a16z's Shift Toward Frontier Tech Investment: AI Infrastructure Development Mirrors Crypto's Past, This Shift May Spur Crypto Talent to Build Frontier Tech
Alliance General Partner Imran recently posted on X, sharing his views on a16z crypto's shift toward frontier technology sectors. He noted that a16z provided early endorsement for the crypto industry when Chris Dixon led Coinbase's investment in 2013; Marc Andreessen's 2014 New York Times op-ed "Why Bitcoin Matters" was a key factor that prompted him to transition into the Bitcoin industry. At the time, very few institutions and funds were willing to publicly support crypto, and a16z and USV were the two that took on brand risk.Imran believes that the current development of AI infrastructure has benefited considerably from Bitcoin mining infrastructure and talent flowing from crypto to AI. Shifting toward frontier tech such as robotics, biology, AI, energy, and compute means that the talent and capabilities accumulated in the crypto industry over the past decade can be used to build the entire frontier tech sector. Citing data centers as an example, he noted that public and government resistance to data centers is growing, much like earlier resistance to crypto—yet today nearly all traditional financial institutions have embraced crypto to some extent.
a16z Crypto rebrands as a16z Frontier, expanding investment scope from crypto to frontier tech startups
a16z co-founder and crypto business lead Chris Dixon announced that a16z Crypto will be renamed a16z Frontier, expanding its investment scope from crypto startups to a broader range of frontier technology sectors, with a focus on supporting companies developing new technologies, creating new markets, and potentially facing longer commercialization cycles. Dixon stated that this adjustment represents business expansion rather than an exit from the crypto space; the fund will continue to invest in crypto and related projects and promote sound crypto regulatory policies. The new fund will also focus on AI and the new products and markets it spawns, including autonomous machines, hardware-software integrated startups, and new types of networks connecting humans and agents.Dixon believes that blockchain will play an important role in this round of technological transformation, that stablecoins are becoming a common tool for fintech companies and traditional institutions, and that blockchain is also expected to support machine-to-machine transactions and provide native economic capabilities for new types of networks. The fund will be jointly led by Dixon, Marc Andreessen, Ben Horowitz, along with Eddy Lazzarin and Guy Wuollet.
Barcelona VC Platform Masia Closes €18 Million First Fund, Will Invest in Early-Stage AI and Frontier Tech Companies
According to EU-Startups, Barcelona venture capital platform Masia has announced the closing of its initial €18 million fund, which will invest in early-stage AI and frontier technology companies tied to Barcelona. The fund has secured backing from over 70 limited partners, including tech founders, corporate executives, and seasoned investors, alongside two institutional investors. At the same time, Masia has launched Masia House in Barcelona’s Gràcia district, a four-story, 1,000-square-meter facility providing office space, meeting rooms, accommodation for visiting investors, and a terrace for events.
Brian Armstrong: Fine-tuned Qwen3.5-9B Outperforms Frontier Large Models in Fraud Prevention Tasks
Odaily reports: Coinbase CEO Brian Armstrong posted on X that after fine-tuning the open-weight model Qwen3.5-9B using Coinbase's internal proprietary data, the model surpassed frontier large models in the specific task of fraud prevention, while significantly reducing costs and running faster.Armstrong believes that in the future, enterprises will increasingly adopt the approach of fine-tuning specialized models for specific tasks, and specialized models are expected to surpass general-purpose frontier models in certain niche tasks. Currently, Coinbase has integrated a model fine-tuning Skill into its internal AI tool framework, allowing employees to initiate the model fine-tuning and deployment process through simple instructions. He also predicts that in the future, some backend services may be completely replaced by this new architecture.
Anthropic Releases Claude Opus 5.5, Cutting Typical Workload Costs by 40%
Anthropic has released Claude Opus 5.5, the first model in the Claude 5.5 series. According to official statements, the model demonstrates improvements in agentic programming, computer use, and knowledge work compared to Opus 5. Under default settings, typical workload costs are reduced by 40%, and output speed increases by over 30%. Additionally, its natural language expression, presentation of key information, and instruction-following capabilities have been enhanced, and the model underwent evaluation by external organizations such as METR and Frontier Design prior to release.
Related news
Sky Q3 net revenue reaches $37.19 million, posting positive growth for six consecutive quarters.
According to disclosures from Sky Ecosystem Insights, the Q3 2026 report released by the Sky Frontier Foundation shows that Sky's net revenue reached $37.19 million in the quarter, representing a year-on-year increase of 11.8% and marking six consecutive quarters of positive growth. The net profit margin rose from 31% in the previous quarter to 37.5%. During the same period, total protocol revenue amounted to $99.29 million, up 14.2% year-on-year, while total expenses decreased by 16% year-on-year to $62.10 million. At quarter-end, the protocol's total collateralized assets reached $11.41 billion, with sUSDS supply standing at $4.52 billion, up 118% year-on-year.
CFTC Chair: AI Agents Could Stake Ethereum and Mint Tokens
According to Cointelegraph, U.S. Commodity Futures Trading Commission (CFTC) Chairman Michael Selig stated at the commission's inaugural Frontier Forum in Washington D.C. that envisioning a refrigerator autonomously purchasing groceries using staked Ethereum proceeds, or even issuing its own memecoin, "is no longer science fiction."
Alliance Partner Discusses a16z's Shift Toward Frontier Tech Investment: AI Infrastructure Development Mirrors Crypto's Past, This Shift May Spur Crypto Talent to Build Frontier Tech
Alliance General Partner Imran recently posted on X, sharing his views on a16z crypto's shift toward frontier technology sectors. He noted that a16z provided early endorsement for the crypto industry when Chris Dixon led Coinbase's investment in 2013; Marc Andreessen's 2014 New York Times op-ed "Why Bitcoin Matters" was a key factor that prompted him to transition into the Bitcoin industry. At the time, very few institutions and funds were willing to publicly support crypto, and a16z and USV were the two that took on brand risk.Imran believes that the current development of AI infrastructure has benefited considerably from Bitcoin mining infrastructure and talent flowing from crypto to AI. Shifting toward frontier tech such as robotics, biology, AI, energy, and compute means that the talent and capabilities accumulated in the crypto industry over the past decade can be used to build the entire frontier tech sector. Citing data centers as an example, he noted that public and government resistance to data centers is growing, much like earlier resistance to crypto—yet today nearly all traditional financial institutions have embraced crypto to some extent.
a16z Crypto rebrands as a16z Frontier, expanding investment scope from crypto to frontier tech startups
a16z co-founder and crypto business lead Chris Dixon announced that a16z Crypto will be renamed a16z Frontier, expanding its investment scope from crypto startups to a broader range of frontier technology sectors, with a focus on supporting companies developing new technologies, creating new markets, and potentially facing longer commercialization cycles. Dixon stated that this adjustment represents business expansion rather than an exit from the crypto space; the fund will continue to invest in crypto and related projects and promote sound crypto regulatory policies. The new fund will also focus on AI and the new products and markets it spawns, including autonomous machines, hardware-software integrated startups, and new types of networks connecting humans and agents.Dixon believes that blockchain will play an important role in this round of technological transformation, that stablecoins are becoming a common tool for fintech companies and traditional institutions, and that blockchain is also expected to support machine-to-machine transactions and provide native economic capabilities for new types of networks. The fund will be jointly led by Dixon, Marc Andreessen, Ben Horowitz, along with Eddy Lazzarin and Guy Wuollet.
Barcelona VC Platform Masia Closes €18 Million First Fund, Will Invest in Early-Stage AI and Frontier Tech Companies
According to EU-Startups, Barcelona venture capital platform Masia has announced the closing of its initial €18 million fund, which will invest in early-stage AI and frontier technology companies tied to Barcelona. The fund has secured backing from over 70 limited partners, including tech founders, corporate executives, and seasoned investors, alongside two institutional investors. At the same time, Masia has launched Masia House in Barcelona’s Gràcia district, a four-story, 1,000-square-meter facility providing office space, meeting rooms, accommodation for visiting investors, and a terrace for events.
Brian Armstrong: Fine-tuned Qwen3.5-9B Outperforms Frontier Large Models in Fraud Prevention Tasks
Odaily reports: Coinbase CEO Brian Armstrong posted on X that after fine-tuning the open-weight model Qwen3.5-9B using Coinbase's internal proprietary data, the model surpassed frontier large models in the specific task of fraud prevention, while significantly reducing costs and running faster.Armstrong believes that in the future, enterprises will increasingly adopt the approach of fine-tuning specialized models for specific tasks, and specialized models are expected to surpass general-purpose frontier models in certain niche tasks. Currently, Coinbase has integrated a model fine-tuning Skill into its internal AI tool framework, allowing employees to initiate the model fine-tuning and deployment process through simple instructions. He also predicts that in the future, some backend services may be completely replaced by this new architecture.